云计算
计算机科学
云测试
服务(商务)
标杆管理
工作量
集合(抽象数据类型)
背景(考古学)
分布式计算
云安全计算
操作系统
生物
经济
业务
古生物学
经济
营销
程序设计语言
标识
DOI:10.1080/08874417.2018.1429957
摘要
In this study, cloud performance is benchmarked in the context of web service using a workload approach with cloud service architecture as a variable. Three sets of cloud service architectures are studied to measure their impact on cloud service performance. Set 1 compares coupled applications against decoupled applications service; Set 2 compares cached applications against non-cached applications service; and Set 3 compares applications with load balancing service against those without. Hypotheses are developed and benchmarking tests conducted to compare the differences in cloud service performance. Among the 3 sets of hypotheses, the statistical analysis results show the first and the third set of architectures have significant differences in cloud service performance, while the second set has no significant difference. The important implications are that cloud performance is a contextual variable; cloud service performance is a cloud architecture driver; and architecture options should be exhausted before other options are considered when cloud service performance needs to be improved.
科研通智能强力驱动
Strongly Powered by AbleSci AI